update paper link + bug fix

examples/mms/lid_rerank/README.md

@@ -3,7 +3,7 @@ This project provides N-best re-ranking, a simple inference procedure, for impro
 
 The workflow is as follows: 1) run LID+ASR inference (MMS and Whisper are supported), 2) compute external re-ranking features, 3) tune feature coefficients on dev set, and 4) apply on test set.
 
-For more information about our method, please refer to the paper: "Improving Multilingual ASR in the Wild Using Simple N-best Re-ranking".
+For more information about our method, please refer to the paper: ["Improving Multilingual ASR in the Wild Using Simple N-best Re-ranking"](https://arxiv.org/abs/2409.18428).
 
 ## 1) Commands to Run LID+ASR Inference
 
@@ -112,4 +112,4 @@ The re-ranked LID and ASR will be in `"path/to/rerank/results"/reranked_1best_li
   journal={arXiv},
   year={2024}
 }
-```
+```

examples/mms/lid_rerank/mms/prep_wav_list.py

@@ -4,6 +4,7 @@
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(description='Example argument parser')
     parser.add_argument('--src', type=str)
+    parser.add_argument('--dst', type=str)
     args = parser.parse_args()
 
     wavs = [x.strip() for x in open(args.src, "r").readlines()]